Acta Biomaterialia xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties David L. Safranski a,b,⇑, Daiana Weiss c, J. Brian Clark c, W. Robert Taylor c,d,e, Ken Gall b,f a
MedShape, Inc., 1575 Northside Drive NW Suite 440, Atlanta, GA 30318, USA School of Materials Science and Engineering, Georgia Institute of Technology, 771 Ferst Drive, Atlanta, GA 30332, USA c Department of Medicine, Division of Cardiology, Emory University School of Medicine, 1639 Pierce Drive, Atlanta, GA 30322, USA d Atlanta Veterans Affairs Medical Center, 1670 Clairmont Rd, Decatur, GA 30033, USA e W.H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive, Atlanta, GA 30332, USA f Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30332, USA b
a r t i c l e
i n f o
Article history: Received 11 November 2013 Received in revised form 30 March 2014 Accepted 15 April 2014 Available online xxxx Keywords: Mechanical properties Glass transition temperature Biodegradation Acrylics Biocompatibility
a b s t r a c t Biodegradable polymers are clinically used in numerous biomedical applications, and classically show a loss of mechanical properties within weeks of implantation. This work demonstrates a new class of semi-degradable polymers that show an increase in mechanical properties through degradation via a controlled shift in a thermal transition. Semi-degradable polymer networks, poly(b-amino ester)-comethyl methacrylate, were formed from a low glass transition temperature crosslinker, poly(b-amino ester), and high glass transition temperature monomer, methyl methacrylate, which degraded in a manner dependent upon the crosslinker chemical structure. In vitro and in vivo degradation revealed changes in mechanical behavior due to the degradation of the crosslinker from the polymer network. This novel polymer system demonstrates a strategy to temporally control the mechanical behavior of polymers and to enhance the initial performance of smart biomedical devices. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Biodegradability is a highly desired property for biomedical polymers used clinically. It allows for tissue growth into the material as the material degrades, enables the release of pharmaceutical agents, and negates the need for device removal [1–6]. Current clinically used biodegradable polymers, such as poly(L-lactide) and poly(caprolactone), take years to degrade, and are associated with some late-stage inflammatory reactions due to incomplete degradation [7–10]. Unfortunately, biodegradable polymers often experience a rapid loss of stiffness and strength upon implantation as the material begins to lose its structural integrity, calling into question their use as a structural support in load-bearing conditions [11–13]. Studies have focused on altering the chemistries of current clinical biodegradable polymers (e.g. by varying the ratio of poly(L-lactide) (PLA)/(poly(lactic-co-glycolic) acid (PLGA)) in an attempt to tailor their degradation rate and the subsequent decrease in mechanical properties [11,14]. PLGA also has even been self-reinforced in another attempt to maintain mechanical
⇑ Corresponding author at: MedShape, Inc., 1575 Northside Drive NW Suite 440, Atlanta, GA 30318, USA. Tel.: +1 404 249 9155; fax: +1 404 249 9158. E-mail address:
[email protected] (D.L. Safranski).
properties, but the loss of mechanical properties occurs nevertheless, albeit at a slower rate, because PLGA’s chemical structure, the driving factor, has not been changed [13]. While the loss of stiffness and strength are inherently inevitable, it is important that the degradation rate be tuned to the rate of tissue infiltration, such that the decrease in mechanical properties occurs in parallel with the gradual shift of structural support provided by the surrounding tissue ingrowth [15–17]. A number of studies have detailed the influence of the glass transition temperature (Tg) on the mechanical properties of (meth)acrylate networks [18–22], especially under physiologically simulated conditions. The mechanical behavior of amorphous polymer networks is highly temperature dependent: a polymer can exhibit glassy or rubbery behavior if cooled below or heated above its Tg, respectively. If heating or cooling is not possible in the desired environment, then immersion in water can be used to shift the Tg, effectively altering the modulus and mechanical behavior. Typically, immersion in water lowers the Tg relative to the testing temperature, which causes the modulus of the material to decrease. This softening of nondegradable plastics is broadly referred to as plasticization; however, this softening does not always lead to a reduction in the toughness, but may actually increase toughness [22]. These mechanical changes occur due to water–polymer
http://dx.doi.org/10.1016/j.actbio.2014.04.022 1742-7061/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
2
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
Fig. 1. (Top) synthesis schematic for PBAE crosslinkers. (Bottom) network formation with PBAE (blue) and MMA (red). As degradation occurs, PBAE leaves thie networks, which increases overall MMA content, increases Tg and changes the mechanical behavior.
interactions and the original thermal properties of the polymer. While these interactions do not degrade the network, they greatly influence the mechanical properties by reducing Tg. When exposed to aqueous conditions, the water–polymer interactions typically decrease Tg through the formation of hydrogen bonds via ‘‘bound’’ water [23–25]. Immersion studies of nondegradable (meth)acrylate networks have shown that these polymers exhibit an initial decrease in Tg followed by an increase in Tg back to the original value over several months. This Tg shift corresponds to changes in the mechanical behavior, including a decrease in moduli from 500 to 15 MPa (i.e. ‘‘plasticization’’) [26]. In addition, the Tg can be predicted via a rule of mixtures, dependent upon each component’s Tg and weight percentage, which allows for facile tailoring of the thermomechanical properties [27–28]. In order for the modulus of a copolymer to increase as it degrades, it can be hypothesized that the Tg should increase as the low-Tg component degrades. A poly(b-amino ester) (PBAE) network containing degradable ester groups in its backbone and acrylate endgroups for photopolymerization can be produced from the step-growth polymerization of a diacrylate and a primary amine [29–31]. The molecular weight, degradation rate and mechanical properties of these networks can be tailored by selecting from a large library of diacrylate and amine combinations. Previous studies have separately examined the combinatorial characterization of PBAE networks and the thermomechanical properties of (meth)acrylates [18,29]. Previous work has also demonstrated that by mixing with varying amounts of methyl methacrylate (MMA), the thermomechanical properties of hydrophobic PBAE networks can be tuned to increase Tg, increase toughness and maintain a rubbery state [32]. However, the relationships between hydrophilicity and thermomechanical behavior have yet to be assessed throughout the degradation process. The objective of this study was to explore the effect of degradation on the temporal mechanical properties of semi-degradable
PBAE–co-MMA networks with varying levels of initial hydrophilicity (Fig. 1). The influence of environment was also examined, where samples were tested in vitro and in vivo to verify that this effect was not limited to an in vitro setting. Three biodegradable crosslinkers, poly(b-amino ester)s with acrylate endgroups, were synthesized with varying levels of hydrophilicity to alter the mass loss rate [31,33]. By combining them with a nondegradable, high Tg monomer, such as methyl methacrylate, the Tg of the network was increased to ambient conditions [32]. The biodegradable crosslinkers have shown previous biocompatibility in vivo, and PMMA is well known to be biocompatible [33–35]. In order for the PBAE–MMA network to offer the same benefits as standard biodegradable polymers, it must elicit a favorable biological response, while also showing an increase in modulus during degradation. The results of this study present (i) how the chemical structure can be tailored to present an increase in mechanical properties with degradation time and (ii) how this process provides a favorable in vivo response using a mouse model.
2. Materials and methods 2.1. Materials and polymer synthesis Poly(ethylene glycol) diacrylate Mn 700 (PEGDA), hexanediol diacrylate (HDDA), 3-methoxypropylamine (3MOPA), methyl methacrylate and 2-hydrodxy-1-[4-(hydroxyethoxy)phenyl]-2methyl-1-propanone (Irgacure 2959) were used as received from Sigma Aldrich. The formation of the semi-degradable network is shown in Fig. 1. Briefly, PEGDA and HDDA were mixed in molar ratios of 0:100, 10:90 and 25:75 to create the diacrylate mixture. This diacrylate mixture was combined with 3MOPA at a molar ratio at 1.15:1 to ensure acrylate endgroups. The step-growth polymerization to form degradable PBAE macromers occurred for 24 h
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
at 200 rpm at 90 °C on a JKEM reaction block (RBC-20 with BTS1500 shaker) following Refs. [31,33]. The resulting macromers were verified by 1H nuclear magnetic resonance spectrometry on a Varian Mercury Vx 400 in deuterated chloroform—the chemical structure can be found in Ref. [33]; the PEGDA units are incorporated between the HDDA units of the macromer. The molecular weight of the PBAE macromers was determined by comparing the number of hydrogens specific to acrylate and amine groups as described in Refs. [31,33]. The average molecular weights of the macromers were 3343, 2602 and 2407 g mol1 for the 25:75, 10:90 and 0:100 macromers, respectively. The PBAE macromers were mixed with MMA (45 wt.% PBAE macromer: 55 wt.% MMA) and 0.5 wt.% Irgacure 2959. This solution was photopolymerized for 45 min under a UVP Blakray lamp (10 mW cm2) [32]. A Bruker Vector 22 Fourier transform IR (FTIR) spectrometer with a Pike Technologies Miracle attenuated total reflectance (ATR) attachment with a ZnSe crystal was used to characterize the chemical structure of the networks after photopolymerization. 2.2. Mass loss and water content 1 cm 1 cm 1 mm samples were weighed, Mo, then soaked in phosphate-buffered saline (PBS), pH 7.4, in an incubator at 37 °C for up to 8 weeks. Samples were removed at time i and their wet mass taken, Mwi. The samples were dried for 24 h at 80 °C and the mass of the samples taken again, Mdi, to determine the mass loss from Eq. (1) and the water content from Eq. (2):
mass loss ¼ 1
Mdi Mo
M wi water content ¼ 1 M di
2.5. In vivo assessment All animal experiments were conducted in accordance with the IACUC guidelines of Emory University. Eight week old, male C57/ BL6 mice were used in the present study. The mice were divided into three groups with networks of 0:100, 10:90 and 25:75 (PEGDA:HDDA molar ratio) networks as well as a control group with no implantation of the polymer network. Mice were anesthetized with Xylazine (13 mg kg1) and Ketamine (87 mg kg1). After creating a small incision in the back of the mouse, a small pocket was made to fit the implant. Animals were sutured closed and given Buprenex (0.01 mg) to alleviate pain. All the mice were on a standard chow diet (Purina, Certified Rodent Chow 5001). Animals were euthanized by CO2 inhalation at 2 and 8 weeks post-surgery. The implants were removed and subjected to mechanical testing. The surrounding tissue was embedded in paraffin, and 5 lm thick serial sections were used for histological analysis. Masson’s trichrome (Sigma) and hematoxylin & eosin were used to visualize collagen/fibrosis and morphology, respectively. To image macrophages immunohistochemical antibodies against Mac3 were used. The primary antibodies were a monoclonal rat anti-mouse macrophage antibody (Mac3) diluted 1:50 (BD Bioscience Pharmingen). Goat anti-rat immunoglobulin diluted 1:200 was used as a secondary antibody. Immunoreactivity was visualized by using streptavidin-conjugated QDots 605 (Invitrogen) diluted 1:100. 2.6. In vivo degradation
ð2Þ
Dynamic mechanical analysis (DMA; TA Instruments Q800) of rectangular samples from in vitro degradation conditions after drying for 24 h was run in tension under strain control, 0.1% strain at 1 Hz following: equilibrate at 100 °C, isothermal hold for 2 min, ramp 3 °C min1 to 200 °C. Tg was defined as the peak of the tan d curve and the rubbery modulus was defined as the storage modulus at a temperature, Tg+75 °C. Crosslinking density, m, was calculated from the rubbery modulus, ER, gas constant, R, and temperature in Kelvin, T, where T = Tg+75 °C, using the following equation:
ER 3RT
required to break the materials and calculated as the area under the stress–strain curve.
ð1Þ
2.3. Thermal properties
m¼
3
ð3Þ
Differential scanning calorimetry (DSC; TA Instruments Q100) was used to determine the glass transition temperature of the samples under both in vitro and in vivo degradation conditions after drying for 24 h. The samples underwent the following treatment: equilibration at 90 °C, isothermal hold for 2 min, then ramp at a rate of 3 °C min1 to 200 °C. Tg was determined from the intersecting line method of the midpoint of the second-order thermal transition. 2.4. Mechanical properties Strain-to-failure tensile tests were performed on a MTS Insight 2 with a 100 N load cell using ASTM D-638 Type IV half-sized dogbone samples at a strain rate of 103 s1 in an environmental chamber filled with PBS and held at 37 °C. For in vitro degradation conditions, dog-bone samples were soaked in PBS in an incubator at 37 °C for up to 8 weeks. Toughness was defined as the energy
ASTM D-638 type IV quarter-sized dog-bone polymer samples were weighed prior to subcutaneous implantation for 2 or 8 weeks. The samples were sterilized under a UV light at 254 nm for 60 min. After extraction, tensile tests were performed, followed by drying for 24 h at 80 °C, weighing to determine mass loss via Eq. (1), followed by FTIR-ATR to determine chemical changes, and thermal characterization by DSC. 2.7. Statistical analysis One-way ANOVA with Tukey’s HSD package in Igor Pro was used to determine if any statistically significant difference was present. p < 0.05. The mean ± standard deviation is given, where n = 3, unless otherwise specified. 3. Results 3.1. Mass loss and water content Two diacrylates, PEGDA and HDDA were mixed in varying molar ratios to control the hydrophilicity of the PBAE crosslinker and resulting PBAE–MMA networks. The mass loss profiles are detailed by showing mass loss over time from in vitro samples, which were immersed in PBS at 37 °C (Fig. 2A) or in vivo samples, where were implanted subcutaneously (Supplementary Fig. 1). For in vitro samples at 1 week, the average mass loss (g/g) of the 25:75 network was 0.11, which is greater than the average mass loss of the 10:90 and 0:100 networks, 0.016 and 0.011, respectively. At 8 weeks, the average mass loss of the 25:75 network was 0.30, while the average mass loss of the 10:90 and 0:100 networks was 0.124 and 0.029, respectively. At 4 weeks, the average mass loss of the 10:90 network, 0.055, had increased above that of the 0:100 network, 0.023. Similar to the in vitro samples, the mass loss of the in vivo samples increased over the 8 weeks. The 25:75
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
4
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
Fig. 2. (A) Mass loss profile of PEGDA:HDDA 1.15 + 55% MMA networks of varying PEGDA:HDDA ratio over 8 weeks under in vitro degradation conditions. (B) Water content for in vitro degraded samples. Each point is the mean ± standard deviation, n = 3.
network’s mass loss was higher than that of the 10:90 and 0:100 networks at 2 weeks and at 8 weeks. In addition, the 10:90 network’s mass loss was higher than the 0:100 at 8 weeks. For the first 2 weeks, the overall water content of the networks increased with time and the water content increased for each network as the PEGDA:HDDA ratio increased. At 2 weeks, water content was 0.436, 0.333 and 0.094 for the 25:75, 10:90 and 0:100 networks, respectively. Beyond 2 weeks, each network’s water content increased with time, but the 10:90 network had a higher water content than the 25:75 network (Fig. 2B). At 8 weeks, water content was 0.6, 0.661 and 0.14 for the 25:75, 10:90 and 0:100 networks, respectively. Hydrolysis of the polymer was also observed by monitoring the carbonyl group and carboxyl group concentration via ATR-FTIR. Over 8 weeks, the carbonyl group concentration decreased and the carboxyl group concentration increased (Supplementary Fig. 2). 3.2. Thermomechanical properties The Tg of each PBAE–MMA network was tracked by DMA for samples degraded under in vitro conditions and dried for 24 h (Fig. 3). DSC was used to track Tg for samples degraded under both in vitro and in vivo conditions and dried for 24 h (Supplementary Fig. 3). DMA and DSC were run after drying in order to eliminate
Fig. 3. Glass transition temperature profile for PEGDA:HDDA 1.15 + 55% MMA networks of varying PEGDA:HDDA ratio over 8 weeks under in vitro degradation conditions determined by DMA. Each point is the mean ± standard deviation, n = 3.
any effects of water on the thermomechanical properties. Two methods of examining Tg were used in order to confirm the increase in Tg, where the DMA values are higher than the DSC values due to differences in how the Tg is defined. The Tg is defined as peak of tan d from DMA, which can be up to 25 °C higher than the DSC values [36–37]. Over 8 weeks, the Tg of each network increased for samples degraded both in vitro and in vivo. As the PEGDA:HDDA ratio increased, the upward shift in Tg increased for each network (e.g. the 25:75 network showed a greater increase in Tg than the 10:90 network). The Tg of the 25:75 network shifted from 48.4 to 95.0 °C, the Tg of the 10:90 network shifted from 50.1 to 72.8 °C, and the Tg of the 0:100 network shifted from 56.5 to 64.2 °C over 8 weeks. By using Eq. (3) and the rubbery modulus of the DMA curves (Fig. 4A), the crosslinking density was calculated. The crosslinking density decreases over 8 weeks (Fig. 4B). As the PEGDA:HDDA ratio increased, the crosslinking density decreased at each time point. Over 8 weeks, the 0:100 network decreased from 309 to 290 mol m3, the 10:90 network decreased from 223 to 131 mol m3, and the 25:75 network decreased from 209 to 81 mol m3. 3.3. Mechanical properties The mechanical behavior of the three PBAE–MMA networks was followed over 8 weeks under both in vitro (Fig. 5) and in vivo (Supplementary Fig. 4) degradation conditions. The 0:100 network maintained rubbery behavior over 8 weeks under both degradation conditions. The 10:90 network initially exhibited rubbery behavior, but by 8 weeks the mechanical behavior had a defined elastic region and plastic region under both in vitro and in vivo conditions. The 25:75 network displayed rubbery behavior at early time points, but by 2 weeks the network displayed defined elastic and plastic regions that lasted through week 6; however, week 8 shows a more brittle response in vitro. The in vivo mechanical behavior of the 25:75 network shows a similar response over the 8 weeks with initial rubbery behavior, followed by defined elastic and plastic regions, then ultimately brittle behavior. The elastic modulus profile for in vitro degradation is presented in Fig. 6 and for in vivo degradation in Supplementary Fig. 5. The 0:100 network’s modulus remained below 5 MPa over the course of 8 weeks with no significant changes. The 10:90 network maintained its modulus over the first 6 weeks in vitro, but showed a significant increase from 6.1 to 18.2 MPa at 8 weeks compared to 6 weeks. In addition, the 10:90 network degraded in vivo maintained a rubbery modulus for the first 2 weeks and showed an increase in modulus at 8 weeks. The 25:75 network displayed the largest increase in modulus, increasing by two orders of magnitude
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
5
Fig. 4. (A) Exemplary DMA curves for the 25:75 PEGDA:HDDA 1.15 + 55% MMA network over 8 weeks of in vitro degradation. (B) Crosslinking density profile for PEGDA:HDDA 1.15 + 55% MMA networks of varying PEGDA:HDDA ratio over 8 weeks of in vitro degradation. Each point is the mean ± standard deviation, n = 3.
Fig. 5. Exemplary stress–strain curves of PEGDA:HDDA 1.15 + 55% MMA networks of varying PEGDA:HDDA ratio over 8 weeks under in vitro degradation conditions. Ratio 0:100 (A), 10:90 (B) 25:75 (C).
within the first 2 weeks in vitro and further increased over the remainder of the 8 week study. The 25:75 network showed a significant increase in modulus at 2 weeks, 115 MPa, compared to the prior time point at 1 week, 13.3 MPa, and a significant increase in modulus at 8 weeks, 234 MPa, compared to the prior time point at 6 weeks, 101.6 MPa. The 25:75 networks degraded in vivo also showed a large increase in modulus during the first 2 weeks and maintained this higher modulus at 8 weeks.
Toughness profiles over 8 weeks for in vitro degradation conditions are shown in Supplementary Fig. 6. All networks showed an increase in toughness during the first week followed by a decrease at 2 weeks, and then changes in toughness that were composition dependent. The toughness of the 0:100 network increased from 2.5 to 4.7 MJ m3 during the first week, then decreased to 2.1 MJ m3 at 2 weeks and maintained toughness to 2.3 MJ m3 at 8 weeks. The toughness of the 10:90 network increased from 0.8 to
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
6
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
5.5 MJ m3 during the first week, then decreased to 1.2 MJ m3 at 2 weeks, followed by an increase to 3.4 MJ m3 at 6 weeks. The toughness of the 25:75 network increased from 1.0 to 7.6 MJ m3 during the first week, then decreased to 4.0 MJ m3 at 2 weeks, followed by an increase to 10.1 MJ m3 at 4 weeks, and then ultimately dropping to 0.6 MJ m3 by 8 weeks. 3.4. In vivo assessment
Fig. 6. Elastic modulus profile for PEGDA:HDDA 1.15 + 55% MMA networks of varying PEGDA:HDDA ratio over 8 weeks under in vitro degradation conditions. Each point is the mean ± standard deviation, n = 3. ⁄P < 0.05 vs. previous timepoint of same network. ^P < 0.05 for 25:75 network vs. 10:90 network and for 25:75 network vs. 0:100 network at given timepoint.
Along with mechanical testing, the inflammatory reaction was examined for each network. The sections show the four layers of the skin: subcutaneous muscle, subcutaneous fat (minimal in the mouse), dermal and epidermal regions. Fig. 7 (2 weeks time point) and Supplementary Fig. 7 (8 weeks time point) suggest a minimal inflammatory response with all three networks compared to the control tissue at the 2 and 8 weeks time points with no difference observed amongst the various compositions. No gross morphological changes were seen in the tissue over the time course, and collagen deposition was minimal, indicating that the material would not be highly encapsulated. At both time points, collagen content and fibrosis were comparable to the control tissue. At 2 weeks, some macrophage infiltration was observed, suggesting an initial
Fig. 7. Photomicrographs of skin from the back of mice after 2 weeks of implantation. Sections were obtained and stained for H&E, trichrome and with a murine macrophagespecific antibody (Mac3).
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
inflammatory response to the presence of the network. However, macrophage infiltration fell to negligible levels 8 weeks after implantation.
4. Discussion The objective of this study was to evaluate the temporal change in mechanical properties of a semi-degradable polymer network under aqueous conditions as a function of the networks’ initial hydrophilicity (i.e. PEGDA concentration). A biodegradable polymer’s standard course of action in the body is to degrade in order to provide drug release or tissue integration; however, a decrease in mechanical properties typically occurs alongside this process. The PBAE–MMA networks presented in this current study represent a novel ‘‘semi-degradable system’’ with mechanical properties that can be controlled during degradation via control of a thermal shift. These networks are defined as ‘‘semi-degradable’’ as only the PBAE crosslinks are expected to undergo hydrolysis and subsequently leave the network, while the nondegradable MMA portion of the network remains intact (Fig. 1). Here we emphasized the control of mass loss by tailoring the network hydrophilicity (via altering the PEGDA:HDDA ratio). Similar to a previous study, networks of PBAE crosslinker with higher PEGDA concentrations had higher rates of mass loss and increased water absorption due to the more hydrophilic nature of PEGDA compared with HDDA [33]. The degradation process of the PBAE crosslinkers from the PBAE–MMA networks can be broken down into three steps: (1) water uptake; (2) hydrolytic cleavage of the PBAE crosslinks; and (3) mass loss of the PBAE degradation products from the network. The rate and amount of water uptake is dependent not only on PEGDA concentration, but also on changes in the crosslinking density. During the first 2 weeks, the water content of the 25:75 network is greater than the 10:90 network because of the higher concentration of PEGDA. During those first 2 weeks, the PEGDA segments undergo hydrolysis, the crosslinking density decreases, and PEGDA leaves the network. At 4 weeks, the 25:75 network has a lower water content than the 10:90 network because more PEGDA has left the 25:75 network than the 10:90 network. Interestingly, the 25:75 network’s water content does not decrease, but continues to increase because the decrease in crosslinking density provides more volume in the network for further water uptake and hydrophilic acrylic acid groups remain in the nondegraded chains with MMA. Crosslinked networks made solely of PBAE crosslinkers degrade by hydrolysis into bis(b-amino acids), diols and poly(acrylic acid) kinetic chains [29]. Since this is a network composed of PBAE crosslinkers and MMA, there would not be pure poly(acrylic acid) kinetic chains. After hydrolysis, the acrylic acid groups would be expected to remain with the MMA chains. The hydrolysis of the PBAE–MMA networks were verified through FTIR by an increase in intensity of carboxyl groups and decrease of carbonyl groups (Supplementary Fig. 2) as well as by a decrease in the crosslinking density determined from DMA (Fig. 4B). The decrease in the carbonyl group is representative of the hydrolytic cleavage of the PBAE crosslinker’s ester groups, which results in the production of the small molecular weight degradation products. The increase in carboxyl group from the FTIR shows that the acrylic acid groups are being produced and remain in the network. Since PEGDA is considered to be more hydrophilic than HDDA, the crosslinks containing more PEGDA are expected to undergo hydrolysis before the crosslinks containing only HDDA. The crosslinking density decreases rapidly within the first 2 weeks for the networks with PEGDA, but then decreases at a slower rate for the remaining time due to the remaining crosslinks with HDDA. The 0:100 network,
7
which contains only crosslinks based upon HDDA, still undergoes slow, gradual hydrolysis as seen in the decrease in its crosslinking density over the 8 weeks due to its hydrolytically cleavable ester bond. Ultimately, the loss of these PBAE crosslinks from the network is shown in the mass loss profiles. There is limited mass loss for the 0:100 network because it is more hydrophobic, has a lower water content and undergoes less hydrolysis than the other networks. Networks created with only HDDA-based PBAE crosslinkers and no MMA undergo very limited mass loss, 20% by 8 weeks, and thus with the added 55 wt.% MMA, the mass loss is further reduced. The 10:90 and 25:75 networks have higher concentrations of PEGDA, thus have higher water contents and undergo more hydrolysis with larger decreases in crosslinking density, which leads to greater mass loss. Mass loss continues throughout the entire 8 weeks, even though most of the hydrolysis occurs in the first 2 weeks most likely due to the hydrolyzed PEGDA segments being in the process of leaving the network with the remaining HDDA segments still undergoing hydrolysis. The increase in Tg observed over the 8 week degradation period can be attributed to the rate at which the PBAE crosslinker left the network. As a binary polymer system, a copolymer network’s Tg can be determined using a rule of mixtures. The increase in Tg can be explained by the PBAE crosslinker component, with a Tg far lower (40 °C) than that of MMA (120 °C), leaving from the network, which leaves the ‘‘high Tg’’ MMA component in the network and thus effectively raises the overall network’s Tg. As the PEGDA:HDDA ratio increased, the magnitude of the Tg shift increased because networks with PEGDA lost mass at a faster rate than those without PEGDA (partially related to the water content of the networks). Drying the samples prior to analysis by DMA or DSC may be a considered a limitation of the study; however, the samples were dried in order to eliminate any shifts in Tg due to water and to observe how the Tg had changed due to mass loss. If wet samples had been run in DMA or DSC, the Tg may have shown a decrease in Tg instead of an increase in Tg at early timepoints. This increase in Tg with degradation time resulted in the thermomechanical behavior of the networks shift from a rubbery to viscoelastic to glassy state at a rate dependent upon network composition. For example, the 25:75 network demonstrated a 46.4 °C increase in Tg over 8 weeks, corresponding to a rapid shift from rubbery to glassy state over the same time course. In stark contrast, the 0:100 network demonstrated a 7.7 °C increase in Tg over the same period, corresponding to minimal changes in stress–strain behavior between 2 and 8 weeks of immersion. Unlike traditional degradable polymers, continued plasticization (decrease in modulus, strength and failure strain) was not observed despite the water content increasing over time, but rather all networks increased in modulus, further suggesting that the Tg–temperature relationship is still driving the mechanical behavior of the networks, even under aqueous conditions. Toughness is an important property of biomaterials that serve a mechanical function in vivo. Toughness broadly reflects the ultimate stress and strain a material can withstand and can be considered an indicator for material durability [26,38]. While most degradable polymer systems lose their toughness as they degrade, the toughness of the PBAE–MMA networks actually exhibit an increase in toughness with degradation time. Previous studies have demonstrated that toughness is maximized when the Tg of a network aligns closely with environmental temperature corresponding to when the network is in its viscoelastic state [22]. Similarly, the 25:75 PBAE–MMA networks exhibited their maximum toughness at a later time point that corresponded to the stress–strain behavior demonstrating large failure strain with a high yield stress. As Tg continues to increase (i.e. the network continues to lose mass), the network’s toughness begins to decrease as more of the
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
8
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx
PBAE departs from the system and the network transitions to a more glassy, brittle state. Comparatively, these PBAE–MMA networks have toughness similar to the initial toughness of nondegradable crosslinked acrylic polymers, 0.14–50 MJ m3, crosslinked polyurethanes, 50.2 MJ m3, swollen chitosan–PVA hydrogels, 1.76 MJ m3, and PLA, 2.13 MJ m3 [22,39–43]. To the authors’ knowledge, there are few comparable studies of degradable polymers that examine the toughness of degradable polymers over the degradation time course. One study found that poly(anhydride ester) fibers only retained 4.8% of their initial toughness, 20 MJ m3, after 24 h of degradation [44]. These PBAE–MMA networks have the potential to be used in a number of translational research areas. The low macrophage infiltration and minimal collagen deposition noted in the histology indicate that the networks elicit a low long-term inflammatory response, regardless of the composition. The similar changes in mechanical behavior after implantation suggest the same relationship between hydrophilicity, degradation and mechanical properties is upheld under in vivo conditions. Furthermore, these results suggest that the network’s ability to shift in Tg and change in modulus does not produce an unfavorable inflammatory response. These PBAE–MMA networks offer the opportunity to explore the effect of increasing stiffness over time on cellular behavior. These networks could be incorporated into orthopedic scaffolds to help tissue regeneration by stiffening once implanted. It has been shown that stiff scaffolds enhance tissue regeneration when compared to soft degradable scaffolds with a modulus less than the native tissue [45]. Due to the acrylate endgroups, in situ photopolymerization would be possible, which would allow for the delivery of drugs into complex spaces in vivo [5]. These polymer networks can also display shape-memory characteristics if their thermomechanical properties are tuned properly, allowing for their use as an implant material in minimally invasive procedures [46]. 5. Conclusion PBAE–MMA networks allow for a novel approach to temporally control the mechanical properties during degradation. The thermomechanical properties are controlled by the change in Tg that shifts as water uptake and subsequent degradation occurs. This work establishes a platform of polymer networks that can be synthesized from a large number of possible combinations of PBAE and (meth)acrylate monomers. Temporal control and the ability to increase the modulus and change the toughness of the polymer network will serve to enhance the design of novel materials for medical implants and tissue engineering applications. Disclosures D.L.S. is an employee of MedShape, Inc. K. Gall is a consultant for MedShape, Inc. Acknowledgement The authors would like to thank NIH for funding SBIR 1R43HL097437. Appendix A. Figures with essential colour discrimination Certain figures in this article, particularly Figs. 1 and 7, are difficult to interpret in black and white. The full colour images can be found in the on-line version, at http://dx.doi.org/10.1016/ j.actbio.2014.04.022.
Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2014.04. 022.
References [1] Yaszemski MJ, Payne RG, Hayes WC, Langer RS, Aufdemorte TB, Mikos AG. The ingrowth of new bone tissue and initial mechanical properties of a degrading polymeric composite scaffold. Tissue Eng 1995;1:41–52. [2] Whang K, Healy KE, Elenz DR, Nam EK, Tsai DC, Thomas CH, et al. Engineering bone regeneration with bioabsorbable scaffolds with novel microarchitecture. Tissue Eng 1999;5:35–51. [3] Sato M, Maeda M, Kurosawa H, Inoue Y, Yamauchi Y, Iwase H. Reconstruction of rabbit Achilles tendon with three bioabsorable materials: histological and biomechanical studies. J Orthop Sci 2000;5:256–67. [4] Freiberg S, Zhu XX. Polymer microspheres for controlled drug release. Int J Pharm 2004;282:1–18. [5] Anseth KS, Metters AT, Bryant SJ, Martens PJ, Elisseeff JH, Bowman CN. In situ forming degradable networks and their application in tissue engineering and drug delivery. J Controlled Release 2002;78:199–209. [6] Frank A, Rath SK, Venkatraman SS. Controlled release from bioerodible polymers: effect of drug type and polymer composition. J Controlled Release 2005;102:333–44. [7] Walton M, Cotton N. Long-term in vivo degradation of poly-L-lactide (PLLA) in bone. J Biomater Appl 2007;21:395–411. [8] Bos RR, Rozema FR, Boering G, Nijenhuis AJ, Pennings AJ, Verwey AB, et al. Degradation of and tissue reaction to biodegradable poly(L-lactide) for use as internal fixation of fractures: a study in rats. Biomaterials 1991;12:32–6. [9] Pistner H, Gutwald R, Ordung R, Reuther J, Muhling J. Poly(L-lactide): a longterm degradation study in vivo. I. Biological results. Biomaterials 1993;14: 671–7. [10] Woodward SC, Brewer PS, Moatamed F, Schindler A, Pitt CG. The intracellular degradation of poly(epsilon-caprolactone). J Biomed Mater Res 1985;19: 437–44. [11] Laitinen O, Tormala P, Taurio R, Skutnabb K, Saarelainen K, Iivonen T, et al. Mechanical properties of biodegradable ligament augmentation device of poly(L-lactide) in vitro and in vivo. Biomaterials 1992;13:1012–6. [12] Matsusue Y, Yamamuro T, Oka M, Shikinami Y, Hyon S, Ikada Y. In vitro and in vivo studies on bioabsorable ultra-high-strength poly(L-lactide) rods. J Biomed Mater Res 1992;26:1553–67. [13] Tormala P, Vasenius J, Vainionpaa S, Laiho J, Pohjonen T, Rokkanen P. Ultrahigh-strength absorbable self-reinforced polyglycolide (SR-PGA) composite rods for internal fixation of bone fractures: in vitro and in vivo study. J Biomed Mater Res 1991;25:1–22. [14] Nair LS, Laurencin CT. Biodegradable polymers as biomaterials. Prog Polym Sci 2007;32:762–98. [15] Muggli DS, Burkoth AK, Anseth KS. Crosslinked polyanhydrides for use in orthopedic applications: degradation behavior and mechanics. J Biomed Mater Res 1999;46:271–8. [16] Anseth KS, Shastri VR, Langer R. Photopolymerizable degradable polyanhydrides with osteocompatibility. Nat Biotechnol 1999;17:156–9. [17] Weiner AA, Shuck DM, Bush JR, Shastri VP. In vitro degradation characteristics of photocrosslinked anhydride systems for bone augmentation applications. Biomaterials 2007;28:5259–70. [18] Safranski DL, Gall K. Effect of chemical structure and crosslinking density on the thermo-mechanical properties and toughness of (meth)acrylate shape memory polymer networks. Polymer 2008;49:4446–55. [19] Yakacki CM, Shandas R, Safranski D, Ortega AM, Sassaman K, Gall K. Strong, tailored, biocompatible shape-memory polymer networks. Adv Funct Mater 2008;18:2428–35. [20] Yakacki CM, Willis S, Luders C, Gall K. Deformation limits in shape-memory polymers. Adv Eng Mater 2008;10:112–9. [21] Ortega AM, Kasprzak SE, Yakacki CM, Diani J, Greenberg AR, Gall K. Structure– property relationships in photopolymerizable polymer networks: effect of composition on the crosslinked structure and resulting thermomechanical properties of a (meth)acrylate-based system. J Appl Polym Sci 2008;110:1559–72. [22] Smith KE, Sawicki S, Hyjek MA, Downey S, Gall K. The effect of the glass transition temperature on the toughness of photopolymerizable (meth)acrylate networks under physiological conditions. Polymer 2009;50:5112–23. [23] Barnes A, Corkhill PH, Tighe BJ. Synthetic hydrogels: 3. Hydroxyalkyl acrylate and methacrylate copolymers: surface and mechanical properties. Polymer 1988;29:2191–202. [24] Corkhill PH, Jolly AM, Ng CO, Tighe BJ. Synthetic hydrogels: 1. Hydroxyalkyl acrylate and methacrylate copolymers–water binding studies. Polymer 1987;28:1758–66. [25] Yang B, Huang WM, Li C, Li L. Effects of moisuture on the thermomechanical properties of a polyurethane shape memory polymer. Polymer 2006;47:1348–56.
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
D.L. Safranski et al. / Acta Biomaterialia xxx (2014) xxx–xxx [26] Smith KE, Trusty P, Wan B, Gall K. Long-term toughness of photopolymerizable (meth)acrylate networks in aqueous environments. Acta Biomater 2011;7:558–67. [27] Fox TG. Influence of diluent and of copolymer composition on the glass temperature of a polymer system. Bull Am Phys Soc 1956;1:123. [28] Belfiore LA. Physical properties of macromolecules. Hoboken, NJ: John Wiley & Sons; 2010. [29] Anderson DG, Tweedie CA, Hossain N, Navarro SM, Brey DM, Van Vliet KJ, et al. A combinatorial library of photocrosslinkable and degradable materials. Adv Mater 2006;18:2614–8. [30] Brey DM, Erickson I, Burdick JA. Influence of macromer molecular weight and chemistry on poly(beta-amino ester) network properties and initial cell interactions. J Biomed Mater Res A 2008;85:731–41. [31] Safranski DL, Lesniewski MA, Caspersen BS, Uriarte VM, Gall K. The effect of chemistry on the polymerization, thermo-mechanical properties and degradation rate of poly(beta-amino ester) networks. Polymer 2010;51: 3130–8. [32] Safranski DL, Crabtree JC, Huq YR, Gall K. Thermo-mechanical properties of semi-degradable poly(beta-amino ester)-co-methyl methacrylate networks under simulated physiological conditions. Polymer 2011;52:4920–7. [33] Safranski DL, Weiss D, Clark JB, Caspersen BS, Taylor WR, Gall K. Effect of poly(ethylene glycol) diacrylate concentration on network properties and in vivo response of poly(beta-amino ester) networks. J Biomed Mater Res A 2011;96:320–9. [34] Jager M, Wilke A. Comprehensive biocompatibility testing of a new PMMA-hA bone cement versus conventional PMMA cement in vitro. J Biomater Sci Polym Ed 2003;14:1283–98. [35] Lu J, Huang Z, Tropiano P, Clouet D’Orval B, Remusat M, Dejou J, et al. Human biological reactions at the interface between bone tissue and polymethylmethacrylate cement. J Mater Sci Mater Med 2002;13:803–9.
9
[36] Menard KP. Dynamic mechanical analysis: a practical introduction. Boca Raton, FL: CRC Press; 1999. [37] Seyler RJ. Assignment of the glass transition temperature. Philadelphia, PA: ASTM; 1994. [38] Briscoe B. Wear of polymers: an essay on fundamental aspects. Tribol Int 1981;14:231–43. [39] Ware T, Hearon K, Lonnecker A, Wooley KL, Maitland DJ, Voit W. Triple-shape memory polymers based on self complementary hydrogen bonding. Macromolecules 2012;45:1062–9. [40] Hearon K, Gall K, Ware T, Maitland DJ, Bearinger JP, Wilson TS. Postpolymerization crosslinked polyurethane shape memory polymers. J Appl Polym Sci 2011;121:144–53. [41] Wang T, Turhan M, Gunasekaran S. Selected properties of pH-sensitive biodegradable chitosan–poly(vinyl alcohol) hydrogel. Polym Int 2004;53: 911–8. [42] Lakhera N, Smith KE, Frick CP. Systematic tailoring of water absorption in photopolymerizable (meth)acrylate networks and its effect on mechanical properties. J Appl Polym Sci 2013;128:1913–21. [43] Meng B, Deng J, Liu Q, Wu Z, Yang W. Transparent and ductile poly(lactic acid)/ poly(butyl acrylate) (PBA) blends: structure and properties. Eur Polym J 2012;48:127–35. [44] Whitaker-Brothers K, Uhrich K. Poly(anhydride-ester) fibers: role of copolymer composition on hydrolytic degradation and mechanical properties. J Biomed Mater Res Part A 2004;70:309–18. [45] Schlichting K, Schell H, Kleemann RU, Schill A, Weiler A, Duda GN, et al. Influence of scaffold stiffness on subchondral bone and subsequent cartilage regeneration in an ovine model of osteochondral defect healing. Am J Sports Med 2008;36:2379–91. [46] Lakhera N, Laursen CM, Safranski DL, Frick CP. Biodegradable thermoset shapememory polymer developed from poly(b-amino ester) networks. J Polym Sci B 2012;50:777–89.
Please cite this article in press as: Safranski DL et al. Semi-degradable poly(b-amino ester) networks with temporally controlled enhancement of mechanical properties. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.04.022
